关于计算机二级的学术报告,【学术报告】2019年4月12日下午肖爱国教授来我院举办系列学术讲座...

报告人：肖爱国(湘潭大学)

报告人简介:肖爱国，湘潭大学数学与计算科学学院二级教授教授、博士生导师。现兼任国防科技数值算法与模拟湖南省国防科技重点实验室主任、科学工程计算与数值仿真湖南省重点实验室副主任、中国仿真学会仿真算法专业委员会主任委员，中国仿真学会、中国数学会计算数学分会理事，期刊《计算数学》、《数值计算与计算机应用》编委等。研究领域为微分方程数值方法，主持国家863课题1项、国家自科基金面上项目4项及省部级科研项目8项。 在《J. Comput. Phys.》、《J. Sci. Comput.》、《Fract. Calc. Appl. Anal.》、《Adv. Comput. Math.》、《BIT》等SCI源刊物上发表论文70多篇，获国家教学成果二等奖、湖南省教学成果一等奖、教育部自然科学二等奖、宝钢教育奖优秀教师奖等。

报告(一)

报告题目：Finite element methods for some fractional diffusion equations

报告摘要：This talk includes the following two parts. In part one, we propose a space–time finite element method (FEM) for the multi-term time–space fractional diffusion equation. First, some fractional derivative spaces, some finite element spaces and their properties are introduced, and a space–time finite element fully discrete scheme is developed. Second, the existence, uniqueness and stability of the scheme are discussed. Third, under the hypothesis about the singularity of the solution neart=0, the convergence is investigated based on the suitable graded time mesh. In part two, we develop an h-p version of the Petrov-Galerkin FEM for 1D fractional differential equation −0Dα x u + Au = f(x) with Dirichlet boundary condition. First, we show the wellposedness of the corresponding weak form. Then by classical hierarchical polynomials as the trial basis functions, we develop a kind of novel test basis functions for the FEM such that the stiffness matrix becomes an identity matrix and the coefficient matrix often has a small condition number. Moreover, we give some properties of the developed test basis functions, and it is shown that the implementation is easier than that of other FEMs.

报告时间:2019年4月12日(星期五)下午2:00-2:45.

报告地点:科技楼南楼602室

报告(二)

报告题目：Asymptotically optimal approximation of some stochastic integralsand its applications

报告摘要：This talk concerns the approximation of some stochastic integrals used in the strong second-order methods for several classes of stochastic differential equations. An explicit construction of the asymptotically optimal approximation (in the meansquare sense) to these stochastic integrals is proposed based on a Karhunen-Loeve expansion of a Wiener process. This asymptotically optimal approximation is more efficient by comparison with the Fourier series approximation by Kloeden and Platen (1992) and the Taylor approximation by Milstein and Tretyakov (2004). In the numerical test part, we replace the stochastic integrals in the strong second-order methods with our corresponding approximations. The numerical results show that those strong second-order methods can perform very well by using our approximation method.

报告时间:2019年4月12日(星期五)下午3:00-3:45.

报告地点:科技楼南楼602室